Marcel Gani wants to make Asia more intelligent

What is the background to Juniper Networks?
The company was founded in 1996 and is still relatively young. The company was founded based on the idea that there was going to be a revolution in the way information was exchanged. The characteristics of the network that would be built, we realized would be very different. In 1996 the networks were built to do multi-protocol translation, where basically an enterprise had various types of hardware in their office and they wanted them to exchange information. Routers were invented to translate this, and evolved in routing information in IP technology. But that was not their primary goal when the equipment was conceived.
So when Juniper was born we recognized that this new public network (the internet) would be based just on IP technology. We realized that it had to be like something where you are used to picking up the phone and getting a dial tone. You would want to do the same thing with that network - the way you exchanged information. You needed to know there would be something at the other end and the message would go through.
The focus from the beginning was on improving the performance of the public network. A significant amount of information had to be carried which necessitated bandwidth and we had to make the software extremely reliable.
The customer base was the service providers, the carriers who were building this new public infrastructure. We worked hand in hand with them to understand what they required. And it was pretty exciting because nothing had been built, so it was virgin territory. Our first product was the M40, which we introduced successfully at the end of 1998 and since then we have gone through a very dramatic growth.
We are based in Sunnyvale, California, but the company has a presence everywhere in the world. In Asia we are in Hong Kong, Korea, China, Japan, Malaysia, Singapore, Taiwan, Sydney, and India.

Who is your biggest rival?
Cisco. It has come at this market from a different angle. One of the differences between us, is that we have equipment which is purpose built for this task - creating the new public infrastructure. The reception we have got in the market has been very good because our equipment fits the needs of the carriers.
From the beginning it was a two horse race, with Cisco being the incumbent because there was nothing else in the market. It had a 100% market share in the "core" of the network. Juniper has been able to make significant gains and we now have 35% of the market.
We saw the market expand rapidly and that allowed us to expand.
A lot of the initial thrust came from the US, but now we're seeing interest everywhere. A lot of PTTs (public telephone companies) show interest in IP technology and move in that direction. In Europe we have won business from France Telecom and Telefonica, and in Asia with Telekom Malaysia and PT Telkom. More traditional telephone companies are seeing the benefits of moving to the IP infrastructure.
The reason this is a very powerful proposition is simple. Today, an operator is using separate networks to carry different types of traffic. So for each application they have a different network, and the cost of this is very high. They have to dig trenches and put fibre into the ground. For carrying voice they have one network, for data another, for private lines another. The goal has always been to consolidate that traffic and migrate it onto a common infrastructure platform. IP is now at a point where it is possible to do voice, create private networks through a technology called VPNs. And if you don't have the cost of maintaining four or five different networks, you make savings. And as operators are coming under more financial pressure, it is critical for them to make savings. That's why even PTTs have started building IP backbones.

So what sort of cost savings would this mean for a company like SingTel?
If you can eliminate one or two networks you can save enormous amounts.

So your proposition to clients like SingTel, is that Juniper Networks is going to save them money, by cutting their costs?
Exactly. It saves cost on the one side. And it also helps them generate additional revenue. An operator may be offering just basic connectivity for $19.95 a month or something. This means it can be hard for the operator to recoup their cost. What we do add is features. For example, a company might opt not to have private lines in its office but conduct all its business over the internet, and create virtual private networks. And if the operator can sell this solution to companies it can enhance their revenue generating capabilities.

What propotion of your revenues come from Asia?
We don't break down our revenues that way. Asia and Europe we classify as 'international' and roughly 36% comes from there and our target is around 40%. I suspect we'll end the year around 35% and be a bit stronger internationally next year.
The pattern of our revenue tends to be pretty lumpy. An operator will decide to buy a significant amount of equipment and start building the network. Then the next quarter they might not buy any more because they are absorbing the equipment they just received.
What we look at is long term trends. We had a very strong, successful first quarter in Asia, and then the second quarter it was slower. We have seen a resurgence in this quarter.

Which Asian countries represent the greatest growth opportunities?
China and Japan. China is undertaking a major network deployment and Japan has very interesting opportunities.
Korea has potential, but has been weaker recently because of the economy in general.

When you talk about China, is it technology leapfrogging straight onto an IP infrastructure?
In Asia in general people are trying to do exactly that. If you look at colour TV around the world, the US doesn't have the best colour TV system in the world because it was first. As people came later, they did it better because they had the opportunity to introduce newer technology.
Asian companies know they have to move to an IP-centric world and they are not encumbered by all the investment they have in previous systems. So they can build networks from scratch along those lines.

The telecom sector has been one of the hardest hit in the slowdown. Presumably your business has been too. Do you see signs of a turnaround or will it be bleak for the foreseeable future?
The company has done well in this difficult time. We have been able to maintain profitability. We had a profit of $32 million in the most recent quarter on revenue of $201 million. We're happy with the way we've managed through this period.
The reason we have done well is that operators are under such an imperative to save costs, that they don't have a choice - because we will save them costs. And improve their business model.
I suspect we'll see consolidation, and some of the weaker service providers will be bought or go out of business.
We have had over a dozen new customer announcements in the last quarter. It positions us extremely well for when the recovery happens. I can't tell you when it will happen, but it will, having been through cycles before. The imperative is that the basic demand for bandwidth keeps growing. People and companies are putting more and more traffic over the network.
At the moment there is too much capacity at the fibre level. But there isn't such a great problem at our level - the router level.
Our equipment is very valuable too, because we are able to deal with a world in which there are thousands of service providers. And to get from one place to the next, traffic is likely to pass through various service providers networks. There is a whole mechanism of settlement and agreements between them, but we can offer solutions that intimately track the traffic and know how much mileage passes through operator Xs fibre versus operator Y.
We introduced Smart IP Services 6-7 months ago, and it does exactly this. The main problem was when there was a bottleneck, there was no solution out there that could process the data fast enough. We need to make the network smarter. We need to figure out where the data is going, but the more time I spend processing something and asking questions about it, the longer it will take. But what our architecture guarantees is that you will continue to operate at the maximum speed of the network, but still be able to interrogate the packets going through the network, and tell how many are coming from which service providers etcetera.
There is another interesting application of Smart IP.When hackers try to bring down a website they flood it with traffic, and it can't handle it. We have known how to fight this technologically. All you have to do is identify a suspect address and drop this traffic. The problem has been that until recently  the only way to do this was with software, as opposed to at the level of the network/ hardware path. This was costing the service providers an enormous amount in terms of performance. So they could do it, but they  would never turn it on, because of the performance cost. Smart IP allows you to filter packets without any degradation of performance. And if you can offer the customer the connectivity but also this security service, that also can enhance added value and added revenue.
We are bringing additional intelligence to the network.

Given Asian telcos have less debt than their European counterparts, does this make Asia a better market for you to sell your products?
I see a significant opportunity in Asia, mostly because the potential users and data growth is significant. The financial health of the customer is very important to us, but the basic driver is the fundamental need for the network itself. And the signs in Asia are that the trend is up, and especially in wireless. That is also an area where we are making forays. In the next two to three years the number of users of wireless data transfer is expected to surpass fixed line data transfer users.